Identifying opportunities to reduce cervical cancer prevention disparities in Western Washington

Cervical cancer screening (CCS) and human papillomavirus (HPV) vaccination are at the heart of the global cervical cancer elimination movement. However, in the United States (US), less than two thirds of age-eligible adolescents are up to date for HPV vaccination, and screening coverage among minorities and immigrants is still low. This dissertation addressed critical knowledge gaps in CCS adherence among East African immigrants and the association of maternal screening adherence with adolescent HPV vaccination. In addition, it predicted the impact of screening by self-sampling on the burden of cervical cancer among East African immigrants in Washington State. The first chapter examined cervical cancer screening adherence and its correlates among 1,664 East African immigrant females aged 25-65 years with ≥1 primary care clinic visit(s) between 2017 and 2018, using electronic health record (EHR) data from University of Washington (UW) Medicine. It also included an analysis of the screening uptake of overdue women who were retrospectively followed for 12 months. Adherence to cervical cancer screening was 63%, and older age, longer duration of care, higher visit frequency, index visit in an obstetrics and gynecology clinic, having an assigned primary care provider, and breast and colorectal cancer screening adherence were associated with higher CCS adherence. Low body mass index was associated with lower screening adherence. Only 9% of overdue women were screened within 12 months of follow-up, and having commercial health insurance compared to having Medicare/Medicaid was associated with higher screening uptake. The second chapter also used EHR data and assessed the association between maternal CCS and adolescent HPV vaccination among adolescent-mother pairs who had ≥1 primary care visit at UW Medicine between 2018 and 2020. It also assessed the associations between maternal breast cancer screening adherence and HPV vaccination, and maternal receipt of a recent wellness visit with HPV vaccination. For the association of maternal CCS and adolescent HPV vaccination, effect modifications by adolescent sex, maternal language interpreter use, and provider characteristics were further evaluated. Of 4,121 identified adolescents, 3,395 (82%) initiated HPV vaccination, and 2,020 (49%) completed the series. CCS adherence and recent maternal wellness visit were associated with higher HPV vaccination. There was no statistically significant association between maternal breast cancer screening and HPV vaccination. The associations of maternal CCS and adolescent HPV vaccination were stronger for male vs. female adolescents, adolescents with a primary care provider with a specialty in Family Practice vs. Pediatrics, and adolescents with the same primary care provider as their mother’s vs. not. The third chapter estimated the impact of screening by self-sampling relative to primary HPV testing via provider collection as the standard of care on cervical cancer incidence and mortality among East African immigrants using a Markov cohort state-transition model for the natural history of high-risk HPV. A hypothetical cohort of 10,000 East African immigrant women aged 25 years was followed until age 80, and outcomes for various scenarios of screening coverage and colposcopy adherence were compared across the screening strategies. Keeping the colposcopy adherence for self-sampling lower than the standard of care based on empirical evidence balanced out the potential gains in screening coverage; cancer incidence and mortalitywere not significantly different across the strategies except when self-sampling coverage increased to 100%. Screening by self-sampling yields lower incidence and mortality of cervical cancer relative to the standard of care when both strategies have equal colposcopy adherence. Overall, cervical cancer screening adherence among East African immigrants was lower than the national average of 80%, and patients with frequent and long-term connections to the health system had higher screening coverage. As our model predicted, self-sampling can be a better alternative for CCS if followed by increased screening coverage and high compliance for follow-up testing. Nevertheless, it can still be an option for under-screened people who could not get screened otherwise. More importantly, since maternal adherence to cervical cancer screening along with attendance to wellness visits had a positive association with HPV vaccination, leveraging healthcare encounters in addition to designing culturally tailored interventions could potentially improve both CCS adherence and adolescent HPV vaccination.